Abstract
Estrogen receptor α (ER) and p53 are critical prognostic indicators in breast cancer. Loss of functional p53 is correlated with poor prognosis, ER negativity, and resistance to antiestrogen treatment. Previously, we found that p53 genotype was correlated with ER expression and response to tamoxifen in mammary tumors arising in mouse mammary tumor virus–Wnt-1 transgenic mice. These results lead us to hypothesize that p53 may regulate ER expression. To test this, MCF-7 cells were treated with doxorubicin or ionizing radiation, both of which stimulated a 5-fold increase in p53 expression. ER expression was also increased 4-fold over a 24-h time frame. In cells treated with small interfering RNA (siRNA) targeting p53, expression of both p53 and ER was significantly reduced (>60%) by 24 h. Induction of ER by DNA-damaging agents was p53 dependent as either ionizing radiation or doxorubicin failed to up-regulate ER after treatment with p53-targeting siRNA. To further investigate whether p53 directly regulates transcription of the ER gene promoter, MCF-7 cells were transiently transfected with a wild-type (WT) p53 expression vector along with a luciferase reporter containing the proximal promoter of ER. In cells transfected with WT p53, transcription from the ER promoter was increased 8-fold. Chromatin immunoprecipitation assays showed that p53 was recruited to the ER promoter along with CARM1, CBP, c-Jun, and Sp1 and that this multifactor complex was formed in a p53-dependent manner. These data show that p53 regulates ER expression through transcriptional control of the ER promoter, accounting for their concordant expression in human breast cancer. [Cancer Res 2009;69(8):3405–14]
Introduction
Estrogen receptor α (ER) plays a critical role in normal breast development and is also involved in the pathogenesis of breast cancer (1, 2). ER expression defines a subset of cancer patients who, in general, have a better prognosis than patients with ER-negative tumors (3). Epidemiologic studies show a strong correlation between ER expression and other positive indicators, such as lower tumor grade and proliferative index. In addition, there is a well-established relationship between ER expression in breast tumors, response to antiestrogen treatment, and disease-free survival (4, 5). In contrast, ER-negative tumors are more likely to recur, develop metastases, and to have markers of poor prognosis, including high tumor grade and proliferative index and inactivating mutations in p53 (4–12).
Mutations in the p53 tumor suppressor gene are frequently associated with human malignancies, including 20% to 40% of breast cancers (7, 8). Stabilization of wild-type (WT) p53 in response to DNA damage promotes cell cycle arrest and apoptosis, whereas mutations in p53 lead to tumorigenesis and multiple drug resistance in breast cancer (13–15). Like many other transcription factors, regulation of gene expression by p53 is mediated both by direct binding to target gene promoters and interaction with other nuclear proteins such as activator protein, CCAAT binding protein (CBP), and Sp1 (16, 17). The diverse nature of p53 target genes shows p53 involvement in various pathways, including cell cycle control (cyclin G and p21), angiogenesis (GD-AIF and thrombospondin), DNA repair (GADD45), growth factor signaling (IGF-BP3), and apoptosis (bax and FAS; ref. 16).
Animal experiments performed in our laboratory showed that p53 genotype correlated with ER expression and response to tamoxifen in mammary tumors arising in mouse mammary tumor virus–Wnt-1 transgenic mice. Compared with tumors in p53 WT (+/+) mice, both ER mRNA and protein were significantly reduced in tumors from p53 (+/−) mice and further reduced in tumors that had a complete loss of p53 due to loss of heterozygosity, suggesting that expression of these two proteins is linked in mammary cancer.1
R. Fuchs-Young, I. Lambertz, J.K.L. Colby, S.H. Shirley, D. Johnston, L.A. Donehower, S.D. Hursting. Evidence for interdependence of ER expression and p53 genotype in mouse mammary tumors that respond to tamoxifen, in submission.
In this study, we report that p53 regulates expression of ER in breast cancer cells. Our results show that ER gene transcription is regulated by p53 binding to the proximal promoter in conjunction with other transcriptional cofactors, including CARM1, CBP, c-Jun, and Sp1. Our results also provide mechanistic insight into clinical data demonstrating the coexpression of WT p53 and ER in human breast cancers (9–12).
Materials and Methods
Tissue culture. MCF-7, MDA-MB-453, and ZR75.1 human breast cancer cells were maintained in phenol red–free DMEM or RPMI supplemented with 5% (v/v) fetal bovine serum (FBS) in 5% CO2 at 37°C. Cells were plated in phenol red–free DMEM/RMPI containing 5% (v/v) dextran-coated charcoal-stripped FBS to ensure steroid-free conditions. At 60% to 80% confluence, cells were treated as indicated in figure legends. Doxorubicin was obtained from Tocris Bioscience.
P53 small interfering RNA constructs and expression vectors. The small interfering (si)RNA constructs used were siGENOME SMARTpool siRNA reagents targeting p53 or negative controls, including scrambled sequence and lamin A–targeting siRNA constructs (Thermo Fisher Scientific). For transfection using a 6-well plate, siRNA (from 20 nmol/L stock) or plasmid DNA (from 1 μg/μL stock) was mixed with Lipofectamine 2000 (Invitrogen) in serum-free DMEM with 1% bovine serum albumin. After incubating for 15 min, solutions were added to the cells. Twenty-four hours after transfection, cells were treated with doxorubicin (0–100 nmol/L) or ionizing radiation (0–16 Gy) as indicated in figure legends. The WT (pC53-SN3) p53 expression vector was a kind gift from Dr. Bert Vogelstein (Johns Hopkins University, Baltimore, MD; ref. 7).
Luciferase assays. MCF-7 or MDA-MB-453 cells were transiently cotransfected with a luciferase reporter construct containing the ER promoter and WT p53 expression vectors. The ERpromLuc vector was a kind gift from Dr. Ronald Weigel (University of Iowa, Iowa City, Iowa; ref. 18) and contained the ER proximal promoter region from −3500 bp to +210 bp relative to the transcription start site. Approximately 5 × 103 cells per well in a 96-well plate were transfected and/or treated as indicated. Luciferase activity was measured using a luciferase assay kit (Promega). The assay was normalized by cotransfection of a pCMV-β-gal plasmid (Clontech) and β-galactosidase activity was measured using a β-gal assay kit (Galacto-Light; Applied Biosystems). The control empty vector used was pGL2 basic from Promega.
Western blot analysis. Cells were washed once with PBS. Modified radioimmunoprecipitation assay buffer [RIPA; 50 mmol/L Tris-HCl (pH 8), 150 mmol/L NaCl, 1 mmol/L EDTA (pH 8), 1% NP40, 0.25% Na-deoxycholate, 1 mg/mL protease inhibitor cocktail (Sigma-Aldrich), 1 mmol/L Na-orthovanadate, and 1 mmol/L NaF] was added to plates and cells scraped to harvest lysates. After a 15-min incubation on ice, cells were centrifuged at 15,000 rpm for 30 min at 4°C. The supernatant was collected and protein content was determined by BCA method (Pierce). Aliquots were run on SDS-PAGE using 10% acrylamide Criterion precast gels (Bio-Rad). Proteins were then electroblotted onto a polyvinylidene difluoride membrane (Pierce). After blocking the membrane with 5% dry milk in PBS with 0.1% Tween 20, blots were probed with an antibody to p53 (DO-1; Santa Cruz Biotechnologies), ER from Zymed (Invitrogen), or glyceraldehyde-3-phosphate dehydrogenase (GAPDH; Abcam). Secondary antibodies used were as follows: horseradish peroxidase–conjugated anti-rabbit IgG (Cell Signaling Technologies) for ER and GAPDH or horseradish peroxidase–conjugated anti-mouse IgG (Santa Cruz Biotechnologies) for p53. Enhanced chemiluminescence was then used to visualize protein bands (Super Signal West Pico; Pierce). The image was captured on an Image Station 440CF (Kodak Digital Science) using Kodak ID v3.6 software (Kodak Scientific Imaging Systems), and protein expression was analyzed using ImageQuant TL software (Amersham Biosciences). Western blots of ER, p53, and GAPDH were quantified by densitometry. Target protein bands were normalized to the GAPDH band in the same sample and then normalized to the untreated control to calculate the fold change.
Real-time quantitative PCR. RNA was extracted from whole cells using the Absolutely RNA kit, following manufacturer's instructions (Agilent Technologies). cDNAs were synthesized from 5 μg of total RNA using random hexamers (Promega) and reverse transcriptase (Invitrogen Corp.). Real-time quantitative PCR was performed using the ABI Prizm 7700 Sequence Detection System (Applied Biosystems). The primers and probes were obtained from Taqman Gene Expression Assays (Applied Biosystems). Samples were run in duplicate on the same plate. Target gene expression was normalized to the reference gene, TATA binding protein, then normalized to the untreated control to determine the fold change.
Chromatin immunoprecipitation assay. Chromatin immunoprecipitation (ChIP) assays were performed essentially as described (19). Briefly, 8 × 106 cells were cross-linked with 1% formaldehyde for 10 min at 25°C and lysed in RIPA buffer. The precleared extract was sonicated to produce 500-bp fragments and split into 2 portions; 1 was used for control input and the other was incubated with antibody and protein A/G beads (Pierce). After IP, beads were washed four times with IP buffer [0.1% SDS, 1% Triton X-100, 2 mmol/L EDTA (pH 8), 150 mmol/L NaCl, and 20 mmol/L Tris-HCl] and eluted in sample elution buffer (1% SDS, 100 mmol/L NaHCO3). For PCR, the p21 promoter was used as a positive control (forward, CCAGCCCTTGGATGGTTT; reverse, GCCTCCTTTCTGTGCCTGA). Primers used for PCR from the ER promoter regions were as follows: −2094 to −1941: forward, CTGCAAAATG CTCCCAAAGT; reverse, TGTTTGGTATGAAAAGGTCACA; −350 to −289: forward, GGGGAGATCTAACAGAAAGAGAGACAA; reverse, CCCTAGATCTGTCTTTCGCG TTTAT, −128 to −40: forward, GGGAGATCTGCCTGGAGTGATGTTTAAG; reverse, TATGAGATCTGGAGACCAGTACTTAAAG. Identical analyses on the human GAPDH, TATA binding protein, and α-actin promoters were included as controls (GAPDH: forward, GTCCACTGGCGTGTTCACCA; reverse-GTGGCAGTGATGGCATGGAC; TATA binding protein: forward, GACCTATGCTCACACTTCTCATGG; reverse, GAACCTGCCCGACCTCACTGAA; Actin, forward-TCGATATCCACGTGACATCCA; reverse, GCAGCATTTTTTTACCCCCTC). For ChIP studies with doxorubicin-treated cells, MCF-7 cells were treated with 50 nmol/L doxorubicin 4 h before ChIP analysis. Antibodies against CARM1, c-Jun, c-Fos, and Sp1 were from Upstate (Millipore). CBP and p53 antibodies were from Santa Cruz Biotechnologies. The HDM2 antibody was obtained from Cell Signaling Technologies, and the RNA polymerase II antibody was obtained from Active Motif. Protein levels of p53 and associated cofactors were determined by immunoblot with antibodies against p53, CARM1, CBP, c-Jun, c-Fos, HDM2, or Sp1 followed by densitometry as described above.
Statistical analysis. All experiments were performed in triplicate. Numerical data are expressed as mean ± SD. Two group comparisons were analyzed by two-sided Student's t test. Multiple group comparisons were analyzed with ANOVA tests. p value of <0.05 was considered significant. Statistical analysis was conducted and graphical representations of the data were plotted using Excel X for Mac.
Results
Ionizing radiation and doxorubicin up-regulate ER expression. To test whether activation of p53 affected ER gene expression, MCF-7 cells were treated with ionizing radiation or doxorubicin, both of which are known to increase expression of p53 by stabilizing the protein (20). Figure 1 shows results obtained after treatment with ionizing radiation (A and B) or doxorubicin (C and D). A 3-fold induction in p53 mRNA was detected at 4 hours after 8 Gy of ionizing radiation, reaching a peak, 8-fold induction by 12 hours and decreasing again by 24 hours (Fig. 1A). Ionizing radiation similarly induced ER mRNA, although with a delayed time frame. Ionizing radiation (8 Gy) stimulated a 3-fold increase in ER mRNA by 8 hours, which reached a maximum 6-fold induction by 12 hours and was maintained through 24 hours (Fig. 1A).
Up-regulation of p53 by ionizing radiation (IR) or doxorubicin (dox) stimulates increased ER expression in MCF-7 cells. A, ionizing radiation treatment increased p53 and ER mRNA as determined by real-time quantitative PCR. *, P < 0.05 for 8 Gy, at indicated times, versus nonirradiated control by Student's t test. B, treatment of MCF-7 cells with ionizing radiation induced time- and dose-dependent increases in p53 protein expression accompanied by corresponding increases in ER protein. *, P < 0.05 for all ionizing radiation doses versus nonirradiated control by ANOVA. C, doxorubicin increased p53 and ER mRNA levels by 0.5 and 4 h, respectively, as determined by real-time quantitative PCR. Message levels of downstream target genes of p53 (p21) and ER (PR, progesterone receptor) were induced with similar kinetics. *, P < 0.05 for target genes at indicated times versus untreated control by ANOVA. D, treatment of MCF-7 cells with doxorubicin induced a time- and dose-dependent increase in p53 protein expression accompanied by corresponding increases in ER. *, P < 0.05 for 10 nmol/L dose versus untreated control; **, P < 0.05 for 50 and 100 nmol/L doses versus both 10 nmol/L and untreated control by ANOVA. Graphical results represent the means of three independent experiments, with representative Western blots shown; error bars indicate standard deviations.
Up-regulation of p53 by ionizing radiation (IR) or doxorubicin (dox) stimulates increased ER expression in MCF-7 cells. A, ionizing radiation treatment increased p53 and ER mRNA as determined by real-time quantitative PCR. *, P < 0.05 for 8 Gy, at indicated times, versus nonirradiated control by Student's t test. B, treatment of MCF-7 cells with ionizing radiation induced time- and dose-dependent increases in p53 protein expression accompanied by corresponding increases in ER protein. *, P < 0.05 for all ionizing radiation doses versus nonirradiated control by ANOVA. C, doxorubicin increased p53 and ER mRNA levels by 0.5 and 4 h, respectively, as determined by real-time quantitative PCR. Message levels of downstream target genes of p53 (p21) and ER (PR, progesterone receptor) were induced with similar kinetics. *, P < 0.05 for target genes at indicated times versus untreated control by ANOVA. D, treatment of MCF-7 cells with doxorubicin induced a time- and dose-dependent increase in p53 protein expression accompanied by corresponding increases in ER. *, P < 0.05 for 10 nmol/L dose versus untreated control; **, P < 0.05 for 50 and 100 nmol/L doses versus both 10 nmol/L and untreated control by ANOVA. Graphical results represent the means of three independent experiments, with representative Western blots shown; error bars indicate standard deviations.
In addition to the effect on p53 mRNA, ionizing radiation stimulated a 6-fold induction in p53 protein over a 12-hour time frame (Fig. 1B) with maximal expression at 8 hours after treatment. Ionizing radiation stimulated a concordant increase in ER protein expression and results showed a 5-fold induction in ER protein over the same time frame with maximal induction at 8 hours. Protein expression in ionizing radiation–treated groups was significantly different from the untreated control (p ≤ 0.01). Although there was a trend toward dose response, the difference between the three doses did not achieve statistical significance. The kinetics of ER induction are consistent with published results for p21, another p53-responsive gene (21–23). Similar results were observed with the chemotherapeutic agent, doxorubicin. P53 mRNA expression was significantly up-regulated by half an hour (p ≤ 0.006) and reached a peak 7-fold induction at 8 hours (p ≤ 0.005) after doxorubicin treatment (Fig. 1C). ER mRNA expression was increased 2-fold by 4 hours and reached a 6-fold induction by 8 hours. Doxorubicin also affected p53 and ER protein expression. Treatment stimulated a 5-fold induction in p53 and a 4-fold induction in ER protein expression compared with vehicle-treated cells (p ≤ 0.03; Fig. 1D). A partial dose response was observed, with the 10 nmol/L treatment group being significantly different from both the 50 and 100 nmol/L groups, but no difference was observed between the 50 and 100 nmol/L groups.
To investigate functional effects of the up-regulation of ER by p53, progesterone receptor was evaluated as a downstream target of ER. Detectable increases in progesterone receptor mRNA were observed by 8 hours after doxorubicin treatment (p ≤ 0.005; Fig. 1C). Expression of p21, a downstream target of p53, was also significantly up-regulated by doxorubicin, starting at half an hour after doxorubicin treatment and reaching a 7-fold peak by 8 hours, mirroring the kinetics of p53 mRNA induction.
These results show that increased p53 mRNA expression preceded changes in ER mRNA levels and suggest that p53 regulates ER expression, thereby contributing to downstream signaling pathways.
Suppression of p53 is associated with down-regulation of ER expression. As our results showed that ER expression was increased after doxorubicin or ionizing radiation, p53-targeting siRNA constructs were transfected into MCF-7 cells to determine if p53 was required for regulation of ER expression. Addition of p53-targeting siRNA resulted in significant decreases in p53 mRNA (Fig. 2A,, top) as well as a reduction in the message level of its downstream target, p21 (Fig. 2A,, middle). ER mRNA expression was also significantly downregulated by p53-targeting siRNA during this timeframe (Fig. 2A , bottom). As expected, transfection of a scrambled sequence siRNA construct did not affect expression of p53, p21, or ER mRNA.
Targeted knockdown of p53 decreases expression of ER in MCF-7 cells. A, real-time quantitative PCR analyses show decreased expression of p53, p21, and ER mRNA 24 h after treatment with p53-targeting siRNA. *, p < 0.05 for siRNA-treated cells versus vehicle-treated or scrambled sequence siRNA–treated controls, as determined by Student's t test. B, Western blot analyses showed down-regulation of both p53 and ER by treatment with 20 to 100 nmol/L p53-targeting siRNA. P53-targeting siRNA constructs inhibited p53 protein expression by ∼60%, which was accompanied by a 50% decrease in ER expression, as determined by densitometry of Western blots. *, p < 0.05 for all siRNA doses versus mock-transfected control by ANOVA. Graphical results represent the means of three independent experiments, with representative Western blots shown; error bars indicate standard deviations.
Targeted knockdown of p53 decreases expression of ER in MCF-7 cells. A, real-time quantitative PCR analyses show decreased expression of p53, p21, and ER mRNA 24 h after treatment with p53-targeting siRNA. *, p < 0.05 for siRNA-treated cells versus vehicle-treated or scrambled sequence siRNA–treated controls, as determined by Student's t test. B, Western blot analyses showed down-regulation of both p53 and ER by treatment with 20 to 100 nmol/L p53-targeting siRNA. P53-targeting siRNA constructs inhibited p53 protein expression by ∼60%, which was accompanied by a 50% decrease in ER expression, as determined by densitometry of Western blots. *, p < 0.05 for all siRNA doses versus mock-transfected control by ANOVA. Graphical results represent the means of three independent experiments, with representative Western blots shown; error bars indicate standard deviations.
Addition of p53-targeting siRNA also decreased p53 protein expression by 50% within 12 hours and 80% by 24 hours, compared with vehicle-treated controls (p ≤ 0.005; Fig. 2B). ER protein expression levels were unchanged at 12 hours but decreased 60% by 24 hours after treatment with the p53 siRNA. Reduced levels of both proteins persisted for 96 hours after siRNA transfection (Fig. 2B). Nontargeting scrambled sequence and lamin A–targeting siRNA did not affect expression of either p53 or ER proteins (data not shown). These results indicate that targeted depletion of p53 resulted in reduced ER expression at both the message and protein levels.
The effects of doxorubicin and ionizing radiation on ER expression are p53 dependent. To further investigate the mechanism of p53 regulation of ER expression, MCF-7 cells were first transfected with p53-targeting siRNA and then exposed to ionizing radiation or doxorubicin. Consistent with the results in Fig. 1, ionizing radiation stimulated a 5-fold induction of p53 protein expression over a 12-hour time frame, which was accompanied by a 4-fold induction in ER protein (Fig. 3A). Similarly, we detected an 8-fold induction of p53 protein at 12 hours and a coincident 7-fold increase in ER expression after doxorubicin treatment (Fig. 3B). Treatment of cells with p53-targeting siRNA reduced ER protein expression by >50%, similar to results reported in Fig. 2. However, when cells were depleted of p53 by siRNA before doxorubicin or ionizing radiation treatment, ER induction was completely blocked, indicating that up-regulation of ER was p53 dependent.
The effects of DNA-damaging agents on ER expression are p53 dependent. Treatment with 20 nmol/L siRNA 24 h before irradiation abrogated the stimulatory effect of 8 Gy ionizing radiation (A) or 50 nmol/L doxorubicin (B) on both p53 and ER protein expression. *, p < 0.05 for ionizing radiation/doxorubicin-treated groups versus untreated control; **, p < 0.05 for p53-targeting siRNA-treated groups versus untreated control by ANOVA. Graphical results represent the means of three independent experiments, with representative Western blots shown, error bars indicate standard deviations.
The effects of DNA-damaging agents on ER expression are p53 dependent. Treatment with 20 nmol/L siRNA 24 h before irradiation abrogated the stimulatory effect of 8 Gy ionizing radiation (A) or 50 nmol/L doxorubicin (B) on both p53 and ER protein expression. *, p < 0.05 for ionizing radiation/doxorubicin-treated groups versus untreated control; **, p < 0.05 for p53-targeting siRNA-treated groups versus untreated control by ANOVA. Graphical results represent the means of three independent experiments, with representative Western blots shown, error bars indicate standard deviations.
Similar results were obtained with ZR75.1 human breast cancer cells, which also express ER and have WT p53. In the ZR75.1 cell line, as in MCF-7s, the up-regulation of ER by doxorubicin was p53-dependent and targeted depletion of p53 with siRNA resulted in decreased ER protein expression (Supplementary Fig. S1).
P53 regulates the ER proximal promoter. We then transfected MCF-7 cells with a luciferase reporter construct containing the proximal promoter of the ER gene (ERpromLuc; ref. 18) with or without the pC53-SN3 WT p53 expression vector (7) to determine whether p53 directly controls transcription of the ER gene promoter. Treatment with 50 nmol/L doxorubicin increased p53 expression (Fig. 1D) and also induced a 3-fold increase in luciferase activity (Fig. 4A) compared with vehicle (p < 0.001). As shown in Fig. 4A, addition of the WT p53 expression vector pC53-SN3 stimulated an 8-fold increase in luciferase expression (p < 0.001), which was further increased to 12-fold by treatment with doxorubicin (p < 0.001 compared with ERpromLuc alone). Experimental controls including the ERpromLuc vector alone or in combination with the empty pGL2 basic vector (Fig. 4A and B) did not induce luciferase expression above basal levels.
Addition of exogenous WT p53 up-regulates ER promoter activity. A, MCF-7 cells were transfected with 2 μg ER promoter–driven luciferase vector (ERpromLuc) either with or without cotransfection of the WT p53 expression vector, pC53-SN3. Cells were treated with 50 nmol/L doxorubicin 24 h after transfection. Luciferase activity was determined 12 h later. The promoter-less pGL2 basic vector was used as a negative control. *, p < 0.05 for designated treatments versus ERpromLuc alone, by Student's t test. Inset, MCF-7 cells were transfected with p53-targeting siRNA 24 h before transfection with the ERpromLuc vector and luciferase expression was measured 24 h later. *, p < 0.05 for p53 siRNA versus vehicle-treated control or scrambled sequence–treated cells by Student's t test. B, MDA-MB-453 cells, which lack both ER and functional p53 expression, were treated as described for A. *, p < 0.05 for designated treatments versus ERpromLuc alone by Student's t test. Graphical results represent the means of three independent experiments, with representative Western blots shown; error bars indicate standard deviations.
Addition of exogenous WT p53 up-regulates ER promoter activity. A, MCF-7 cells were transfected with 2 μg ER promoter–driven luciferase vector (ERpromLuc) either with or without cotransfection of the WT p53 expression vector, pC53-SN3. Cells were treated with 50 nmol/L doxorubicin 24 h after transfection. Luciferase activity was determined 12 h later. The promoter-less pGL2 basic vector was used as a negative control. *, p < 0.05 for designated treatments versus ERpromLuc alone, by Student's t test. Inset, MCF-7 cells were transfected with p53-targeting siRNA 24 h before transfection with the ERpromLuc vector and luciferase expression was measured 24 h later. *, p < 0.05 for p53 siRNA versus vehicle-treated control or scrambled sequence–treated cells by Student's t test. B, MDA-MB-453 cells, which lack both ER and functional p53 expression, were treated as described for A. *, p < 0.05 for designated treatments versus ERpromLuc alone by Student's t test. Graphical results represent the means of three independent experiments, with representative Western blots shown; error bars indicate standard deviations.
To determine if p53 could regulate not only inducible activity of the ER promoter, but also basal transcription, the p53 targeting siRNA was transfected into untreated MCF-7 cells. Addition of p53-targeting siRNA resulted in a 48% (p = 0.009) decrease in basal expression of the ERpromLuc construct (Fig. 4A,, inset). These data support our earlier results showing that ER expression was decreased after siRNA-mediated knockdown of p53 (Fig. 3) and show that p53 contributes to basal as well as doxorubicin-inducible ER expression.
To better define the effect of p53 on transcriptional activity of the ER gene promoter, experiments were also conducted in the p53-null MDA-MB-453 breast cancer cell line. In the absence of endogenous p53, doxorubicin treatment was unable to induce ER promoter activity in these cells (Fig. 4B). However, addition of WT p53 (pC53-SN3) stimulated a 5-fold increase in luciferase activity, which was enhanced to 14-fold by doxorubicin, compared with the ERpromLuc vector alone. These results show that p53 regulated ER expression in breast cancer cells by increasing transcriptional activity at the proximal promoter.
P53 binds to the ER promoter. To investigate the mechanism whereby p53 regulates the ER gene promoter activity, ChIP assays were used. A recent report indicates that the sequences between −350 bp to −289 bp and −128 bp to −40 bp upstream of the transcriptional start site are involved in transcriptional activity of the ER gene (24). Accordingly, three promoter regions were investigated in these experiments, including the previously identified −350 to −289 and −128 to −40 bp sequences, as well as a further upstream sequence spanning from −2094 to −1941 bp. Because our results showed that doxorubicin treatment up-regulated ER mRNA in MCF-7 cells (Fig. 1D), we reasoned that this treatment would stimulate p53 recruitment to the ER promoter. ChIP assays using a p53 antibody showed that p53 was recruited to the −128 to −40 bp region of the proximal promoter in response to doxorubicin (Fig. 5A). P53 was also detected at the distal regions of the ER promoter (−2094 to −1941 bp and −350 to −289 bp), but these associations were not altered by doxorubicin treatment. As anticipated, after exposure to doxorubicin, p53 was also immunoprecipitated at the p21 promoter but not at the α-actin or TATA binding protein gene promoters.
p53 recruitment to the proximal ER promoter is enhanced after doxorubicin treatment. A, p53 associated with the ER promoter in three regions, but recruitment to the region −128 to −40 bp upstream of the transcriptional start site is enhanced by 50 nmol/L doxorubicin treatment. Recruitment to the promoter region for p21 was used as a positive control and promoters for α-actin and TATA binding protein (TBP) genes were used as negative control regions. B, in ChIP assays, p53 coimmunoprecipitated with CARM1, CBP, c-Jun, c-Fos, HDM2, and Sp1. C, ChIP assays using CARM1, Sp1, c-Jun, and CBP antibodies verified that each of these cofactors was also recruited to the −128 to −40 bp region of the ER promoter after doxorubicin treatment. Pretreatment of MCF-7 cells with 20 nmol/L p53-targeting siRNA, but not nontargeting siRNA, blocked association of each of these cofactors to the −128 to −40 bp region of the ER promoter in response to 50 nmol/L doxorubicin (dox/siRNA versus dox lanes). D, ChIP assays using a RNA pol II antibody verified that this component of the basal transcription machinery was recruited to the −128 to −40 bp region of the ER promoter after doxorubicin treatment. IB, immunoblot.
p53 recruitment to the proximal ER promoter is enhanced after doxorubicin treatment. A, p53 associated with the ER promoter in three regions, but recruitment to the region −128 to −40 bp upstream of the transcriptional start site is enhanced by 50 nmol/L doxorubicin treatment. Recruitment to the promoter region for p21 was used as a positive control and promoters for α-actin and TATA binding protein (TBP) genes were used as negative control regions. B, in ChIP assays, p53 coimmunoprecipitated with CARM1, CBP, c-Jun, c-Fos, HDM2, and Sp1. C, ChIP assays using CARM1, Sp1, c-Jun, and CBP antibodies verified that each of these cofactors was also recruited to the −128 to −40 bp region of the ER promoter after doxorubicin treatment. Pretreatment of MCF-7 cells with 20 nmol/L p53-targeting siRNA, but not nontargeting siRNA, blocked association of each of these cofactors to the −128 to −40 bp region of the ER promoter in response to 50 nmol/L doxorubicin (dox/siRNA versus dox lanes). D, ChIP assays using a RNA pol II antibody verified that this component of the basal transcription machinery was recruited to the −128 to −40 bp region of the ER promoter after doxorubicin treatment. IB, immunoblot.
To identify additional members of the ER-regulating transcriptional complex, coimmunoprecipitations were conducted using antibodies that target several potential cofactors. It was recently reported that CARM-1, CBP, and p53 coimmunoprecipitate in glutathione S-transferase pull-down assays and cooperatively bind at the GADD45 gene promoter (25). Furthermore, in silico analyses identified activating transcription factor (AP-1), CBP, and Sp1 response elements within the ER promoter regions described above (Fig. 6). As shown in Fig. 5B, CARM1, CBP, c-Jun, and Sp1 were detected in complexes immunoprecipitated with p53 antibody after doxorubicin treatment. In contrast, the amount of HDM2 protein coimmunoprecipated with p53 decreased after doxorubicin treatment, consistent with its role as a negative regulator of p53 expression (26, 27). Interestingly, c-Fos did not coimmunoprecipitate with p53 after doxorubicin treatment, indicating that c-Jun was associated with the p53 either as a homodimer or with AP-1 proteins other than c-Fos, perhaps with other members of the Jun, Fos, AP-1, or JDP families.
Schematic representation of the ER promoter. Bent arrow, the transcriptional start site. Shaded boxes represent regions targeted by primer sets. A, AP-1 response elements; C, CBP response elements; S, Sp1 response elements; and T, TATA box sequences. Transcription factor consensus sequences were identified using the web-based Transcription Element Search System program (24, 42, 43).
Schematic representation of the ER promoter. Bent arrow, the transcriptional start site. Shaded boxes represent regions targeted by primer sets. A, AP-1 response elements; C, CBP response elements; S, Sp1 response elements; and T, TATA box sequences. Transcription factor consensus sequences were identified using the web-based Transcription Element Search System program (24, 42, 43).
ChIP assays, using antibodies directed against CARM1, CBP, c-Jun, and Sp1, revealed that these four proteins were also recruited to the −128 to −40 bp region of the ER promoter after treatment with doxorubicin (Fig. 5C). Like p53, binding of all of these cofactors with upstream promoter regions of ER (−2094 to −1941 and −350 to −289) was not altered by doxorubicin treatment (data not shown).
To determine whether p53 was required for binding of CARM1, CBP, c-Jun, and Sp1 to the ER promoter, MCF-7 cells were treated with p53-targeting siRNA constructs followed by ChIP analysis. As shown in Fig. 5C, doxorubicin did not stimulate p53/CARM1/CBP/c-Jun/Sp1 complex formation or DNA binding in cells transfected with p53 siRNA, indicating that p53 was required for assembly of the complex on the −128 to −40 bp region of the ER promoter. To further demonstrate that the −128 to −40 bp region of the ER gene promoter is critical for p53-mediated induction of transcriptional activity, the association of RNA polymerase II was assessed. Pol II binding to the −128 to −40 promoter was enhanced after doxorubicin treatment (Fig. 5D , Dox lane) and, as expected, pol II was not recruited to the upstream regions (−350 to −289; −2094 to −1941) of the promoter after doxorubicin treatment. These results further show that the −128 to −40 bp region is involved in doxorubicin-induced, p53-mediated transcription of the ER promoter.
Discussion
In this study, up-regulation of p53 by DNA-damaging agents, such as ionizing radiation or doxorubicin, resulted in increased ER mRNA and protein expression, whereas siRNA-mediated p53 knockdown reduced ER expression. P53 was shown to directly regulate the ER promoter, as both endogenous as well as transiently expressed p53 increased transcriptional activation of the ER gene promoter in luciferase assays. Furthermore, p53 was recruited to the ER gene promoter as a part of a regulatory complex that included CARM1, CBP, c-Jun, and Sp1. Our analysis of the ER promoter showed that although there are response elements for Sp1, AP-1, and CBP, no p53 consensus response element was present in the proximal promoter, suggesting that p53 regulates the ER gene through protein-protein interactions (24). Furthermore, our data demonstrate that p53 was required for assembly of this complex as well as for induction of ER expression by doxorubicin.
There is accumulating evidence of crosstalk between growth promoting and growth prohibiting pathways, such as those demonstrated here involving ER and p53. Evidence indicates that ER signaling increases expression and nuclear accumulation of p53 in vitro and in vivo (28–30). Recapitulation of the hormonal milieu of pregnancy by exposing rats to estrogen plus progesterone induces nuclear accumulation of p53 in mammary epithelium and protects against carcinogen challenge (31–33). Although this literature represents substantial evidence that ER regulates p53 expression, the converse regulation of ER by p53 has not been studied and a mechanism linking ER expression with WT p53 is virtually unexplored. Although clinical and epidemiologic studies have shown that ER-positive human breast cancers express WT p53 and tumors with inactivating mutations in p53 tend to be ER negative, evidence for a direct transcriptional mechanism has been lacking. Results presented here show that p53 regulates the ER promoter in breast cancer cells and provide a possible explanation for the epidemiologic findings that ER expression is associated with WT p53 in human breast cancers.
P53 is a widely recognized tumor suppressor gene, which limits proliferation by inducing cell cycle arrest or apoptosis (34, 35). Intriguingly, recent reports suggest that p53 may also play a role in regulating genes involved in promotion of cell growth. Specifically, p53 mediates induction of cyclooxygenase-2, transforming growth factor-α, and heparin-binding epidermal growth factor–like growth factor, all of which stimulate growth of normal and neoplastic cells (36–38). Furthermore, it has recently been reported that ER and p53 are associated on the promoter of vascular endothelial growth factor receptor to cooperatively enhance transcriptional activation (39). Consistent with these reports, our data suggest the presence of a functional link between the tumor suppressor p53 and growth-promoting ER signaling pathways.
The regulation of ER transcription is complex, involving multiple independent promoters and several alternative 5′ exons. Although it has previously been shown that Sp1 (40) and GATA-3 (41) are critical regulators of ER expression, the role of p53 in mediating expression of ER remains virtually unstudied. Our results show that despite the lack of a p53 consensus sequence, p53 bound to the proximal ER gene promoter along with CARM1, CBP, c-Jun, and Sp1. Dox stimulated the p53-dependent assembly of the transcription complex at the −128 bp to −40 bp region of the ER proximal promoter but not at the more distal regions. When considered along with the data showing that p53 siRNA reduced luciferase in untreated MCF-7 cells, these results suggest that the −2094 to −1941 and −350 to −289 regions may be responsible for basal transcription from the ER promoter, whereas the −128 to −40 bp region is involved in p53-inducible activity. It has previously been reported that transcription from this proximal promoter region [termed “promoter A” by Kos and colleagues (42) and including the −128 to −40 bp region studied here] predominates in ER-positive breast cancer cells (43, 44), but little is known regarding why this promoter region is preferentially used. The results presented here indicate that p53 plays a role in use of promoter A in ER-positive breast cancers, potentially contributing to the reported overexpression of ER in postmenopausal breast cancer compared with normal mammary tissue (18, 43, 45).
One of the unexplained features of breast tumors with mutations in p53 is that they frequently lack ER expression, which is correlated with lack of response to tamoxifen and poor prognosis (9–12, 46). This is noteworthy as the molecular mechanisms responsible for the ER-negative phenotype are not well-understood but may include progressive loss of ER. Consistent with this is the observation that in transgenic mice with BRCA1 mutations, early mammary lesions are ER positive, but expression is lost during tumor progression coincident with loss of functional p53 (47, 48). Similarly, in human carriers of BRCA1/2 mutations, breast tumors tend to be ER negative, have mutated p53, and have a poorer prognosis than tumors with WT BRCA (49).
Our studies have identified a transcriptional mechanism that explains the observed correlation between p53 function and ER expression in human breast cancers. In this report, we demonstrate that p53 regulates ER expression, providing a mechanism to explain the concordance of WT p53 and ER expression and the high frequency of ER negativity observed in tumors with p53 mutations. Furthermore, these findings suggest that early mutations and/or loss of functional p53 during tumorigenesis may lead to ER-negative breast cancers, which has consequences on tumor progression and response to antihormonal therapies.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
Acknowledgments
Grant support: Susan G. Komen Breast Cancer Foundation BCTR0201390 (R.F. Young), Ruth L. Kirschstein National Research Service Award/Departmental Training grant CA09480 (S.H. Shirley), National Institute of Environmental Health Sciences Center for Research on Environmental Disease (ES07784), and National Cancer Institute Cancer Center Support (CA16672).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Dr. Bert Vogelstein for providing the p53 expression vectors and Dr. Ronald Weigel for the ER promoter luciferase construct, as well as Dr. S.S. Lange for critical reading of the manuscript.